feat: add scalattice openai-compatible provider - #35221
Conversation
| "scalattice": { | ||
| "base_url": "https://api.scalattice.cloud/v1", | ||
| "api_key_env": "SCALATTICE_API_KEY", | ||
| "api_base_env": "SCALATTICE_API_BASE", | ||
| "supported_endpoints": ["/v1/chat/completions"] | ||
| }, |
There was a problem hiding this comment.
Provider coverage metadata is missing
When the provider documentation coverage check loads this new registry key, scalattice has no matching entry in provider_endpoints_support.json, causing the check to raise UndocumentedProviderError.
Rule Used: Block new LLM Provider integrations if they're ope... (source)
Knowledge Base Used: LLM Provider Adapters
Greptile SummaryAdds Scalattice as a JSON-configured OpenAI-compatible provider.
Confidence Score: 4/5The provider coverage failure must be fixed before merging by adding Scalattice to provider_endpoints_support.json. The runtime JSON-provider resolution path handles Scalattice, but the new registry key deterministically violates the repository's provider documentation coverage check and raises UndocumentedProviderError. Files Needing Attention: litellm/llms/openai_like/providers.json and provider_endpoints_support.json
|
| Filename | Overview |
|---|---|
| litellm/llms/openai_like/providers.json | Adds the Scalattice registry entry, but omits the coordinated endpoint-support metadata required by the provider coverage check. |
Reviews (1): Last reviewed commit: "Add Scalattice as an OpenAI-compatible p..." | Re-trigger Greptile
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
Register Scalattice in the openai_like providers registry and document endpoint coverage for chat completions.
144698c to
34c492e
Compare
Register Scalattice in the dashboard provider map with the brand mark, and publish catalog model costs so spend tracking and docs pricing stay aligned.
|
Follow-up pushed on this branch:
Docs page PR (separate repo): BerriAI/litellm-docs#719 |
Rebuild pricing files from upstream staging and re-add only Scalattice catalog entries so backup matches main after merge.
|
Fixed pricing backup pollution: rebuilt model prices from upstream staging and re-added only Scalattice catalog entries so gpt-5.4-mini/nano max_input_tokens stay at 272000 in both files. |
Summary
litellm/llms/openai_like/providers.jsonprovider_endpoints_support.json(chat completions)Providersenum + logo (scalattice.svg)model_prices_and_context_window.json(backup synced)https://api.scalattice.cloud/v1SCALATTICE_API_KEY(optional override viaSCALATTICE_API_BASE)Docs page (separate repo): BerriAI/litellm-docs#719
Usage
Docs: https://scalattice.cloud/docs/developers#litellm
Product: https://scalattice.com/
Test plan
check_provider_folders_documented.pypasseslitellm_internal_stagingscalatticeproviderlitellm.completion(model="scalattice/<model>", ...)withSCALATTICE_API_KEY